Several studies have raised concern regarding the possible association between proton-pump inhibitors (PPIs) andClostridium difficileinfection (CDI). We aimed to perform a systematic review of incident and recurrent CDI in PPI users, and to evaluate the relative impact of concurrent antibiotic use, or switching acid suppression to histamine-2-receptor antagonists (H2RAs).
We searched MEDLINE and EMBASE from inception to December 2011 for controlled observational studies that reported on the risk of CDI with and without PPI use. We performed random effects meta-analysis and assessed statistical heterogeneity using theI2 statistic.
We included 42 observational studies (30 case-control, 12 cohort) totalling 313,000 participants overall. Pooled analysis of 39 studies showed a statistically significant association between PPI use and risk of developing CDI, odds ratio (OR) 1.74 (95% confidence interval (CI) 1.47–2.85,P<0.001,I2=85%) compared with non-users. A pooled analysis of three studies showed a significant associated risk of recurrent CDI associated with PPIs, OR 2.51 (95% CI 1.16–5.44,P=0.005,I2=78%). Subgroup analysis failed to fully clarify the source of the substantial statistical heterogeneity. Adjusted indirect comparison demonstrated that use of H2RAs as an alternative carried a lower-risk OR 0.71 (95% CI 0.53–0.97) compared with PPIs. Conversely, concomitant use of PPI and antibiotics conferred a greater-risk OR 1.96 (95% CI 1.03–3.70) above that of PPIs alone. For PPI and antibiotics, the Rothman's synergy index was 1.36 and attributable proportion of risk from interaction 0.19, indicating an increased risk from interaction beyond the effects of each drug alone.
Despite the substantial statistical and clinical heterogeneity, our findings indicate a probable association between PPI use and incident and recurrent CDI. This risk is further increased by concomitant use of antibiotics and PPI, whereas H2RAs may be less harmful.
1 Norwich Medical School, University of East Anglia, Norwich, UK
2 Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
3 Department of Medicine, Division of Pulmonary, Critical Care and Sleep Disorders Medicine, University of Louisville, Louisville, Kentucky, USA
Correspondence: Yoon Kong Loke, MBBS, MD, Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ, UK. E-mail: firstname.lastname@example.org
SUPPLEMENTARY MATERIAL accompanies this paper at http://links.lww.com/AJG/A483
Received 8 February 2012; accepted 15 March 2012
published online 24 April 2012